@inproceedings{lin-etal-2026-randomization,
title = "A Randomization-Control Diagnostic for Document Grounding in Self-Distillation Gates",
author = "Lin, Xiankun and
Liu, Lixin and
Lin, Xinyi and
Hao, Zhezheng",
editor = "Wang, Yimu and
Choi, Yee Man and
Wu, Di and
Ouyang, Siqi and
Azadani, Mozhgan Nasr and
Fung, Yi R. (May)",
booktitle = "Proceedings of the 1st Workshop on Grounding Language Models: Learning Faithfully and Efficiently ({G}round{LM} 2026)",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.groundlm-1.32/",
pages = "412--417",
ISBN = "979-8-89176-478-1",
abstract = "Self-distillation for grounded question answering often filters tutor trajectories with agreement or confidence gates. For such gates to support faithful grounding, their signals should depend on the supplied document rather than on answers recoverable from parametric memory. We propose a four-mode randomization-control diagnostic that applies the same frozen gate to original, randomized, irrelevant, and empty documents for the same question. On HotpotQA and MuSiQue, using Qwen2.5-7B-Instruct with eight tutor traces, the joint gate changes sharply with document condition: kept fraction falls from 0.380 to 0.020 and from 0.330 to 0.015, respectively. Randomized and irrelevant documents lie between the endpoints, while a document-blind random gate stays flat. The protocol audits document sensitivity without additional student training; faithful evidence use and downstream gains require separate validation."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="lin-etal-2026-randomization">
<titleInfo>
<title>A Randomization-Control Diagnostic for Document Grounding in Self-Distillation Gates</title>
</titleInfo>
<name type="personal">
<namePart type="given">Xiankun</namePart>
<namePart type="family">Lin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lixin</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xinyi</namePart>
<namePart type="family">Lin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhezheng</namePart>
<namePart type="family">Hao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 1st Workshop on Grounding Language Models: Learning Faithfully and Efficiently (GroundLM 2026)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yimu</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yee</namePart>
<namePart type="given">Man</namePart>
<namePart type="family">Choi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Di</namePart>
<namePart type="family">Wu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Siqi</namePart>
<namePart type="family">Ouyang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mozhgan</namePart>
<namePart type="given">Nasr</namePart>
<namePart type="family">Azadani</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yi</namePart>
<namePart type="given">R</namePart>
<namePart type="given">(May)</namePart>
<namePart type="family">Fung</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-478-1</identifier>
</relatedItem>
<abstract>Self-distillation for grounded question answering often filters tutor trajectories with agreement or confidence gates. For such gates to support faithful grounding, their signals should depend on the supplied document rather than on answers recoverable from parametric memory. We propose a four-mode randomization-control diagnostic that applies the same frozen gate to original, randomized, irrelevant, and empty documents for the same question. On HotpotQA and MuSiQue, using Qwen2.5-7B-Instruct with eight tutor traces, the joint gate changes sharply with document condition: kept fraction falls from 0.380 to 0.020 and from 0.330 to 0.015, respectively. Randomized and irrelevant documents lie between the endpoints, while a document-blind random gate stays flat. The protocol audits document sensitivity without additional student training; faithful evidence use and downstream gains require separate validation.</abstract>
<identifier type="citekey">lin-etal-2026-randomization</identifier>
<location>
<url>https://aclanthology.org/2026.groundlm-1.32/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>412</start>
<end>417</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T A Randomization-Control Diagnostic for Document Grounding in Self-Distillation Gates
%A Lin, Xiankun
%A Liu, Lixin
%A Lin, Xinyi
%A Hao, Zhezheng
%Y Wang, Yimu
%Y Choi, Yee Man
%Y Wu, Di
%Y Ouyang, Siqi
%Y Azadani, Mozhgan Nasr
%Y Fung, Yi R. (May)
%S Proceedings of the 1st Workshop on Grounding Language Models: Learning Faithfully and Efficiently (GroundLM 2026)
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-478-1
%F lin-etal-2026-randomization
%X Self-distillation for grounded question answering often filters tutor trajectories with agreement or confidence gates. For such gates to support faithful grounding, their signals should depend on the supplied document rather than on answers recoverable from parametric memory. We propose a four-mode randomization-control diagnostic that applies the same frozen gate to original, randomized, irrelevant, and empty documents for the same question. On HotpotQA and MuSiQue, using Qwen2.5-7B-Instruct with eight tutor traces, the joint gate changes sharply with document condition: kept fraction falls from 0.380 to 0.020 and from 0.330 to 0.015, respectively. Randomized and irrelevant documents lie between the endpoints, while a document-blind random gate stays flat. The protocol audits document sensitivity without additional student training; faithful evidence use and downstream gains require separate validation.
%U https://aclanthology.org/2026.groundlm-1.32/
%P 412-417
Markdown (Informal)
[A Randomization-Control Diagnostic for Document Grounding in Self-Distillation Gates](https://aclanthology.org/2026.groundlm-1.32/) (Lin et al., GroundLM 2026)
ACL